#1549 Alessandro De Carli | Is Bitcoin Decentralization Coming To Your Phone?!
🎯 Summary
Podcast Episode Summary: #1549 Alessandro De Carli | Is Bitcoin Decentralization Coming To Your Phone?!
This episode of the Pom Podcast features Anthony Pompliano interviewing Alessandro Di Carli, co-founder of Accurus, about their ambitious project to build a decentralized computational network utilizing the latent power of everyday mobile phones. The conversation draws strong parallels between the success and resilience of the Bitcoin network and the potential for decentralized general compute.
1. Focus Area
The primary focus is on Decentralized Infrastructure and Compute, specifically exploring how the principles of decentralization, proven by Bitcoin, can be applied to harness the vast, underutilized computational power residing in mobile devices to create a resilient, globally accessible compute layer. The discussion heavily touches upon mobile security (hardware security modules/TEEs) as the foundational element for this new compute network.
2. Key Technical Insights
- Mobile Hardware Security as Foundation: Mobile devices are uniquely positioned for decentralized compute because they possess advanced Hardware Security Modules (HSMs), enabling critical properties like verifiability (proving work was done correctly) and confidentiality (preventing the host from inspecting data).
- Compute Concentration Misconception: The vast majority of global compute power is not in data centers but embedded within household and pocket-sized mobile devices, which are currently largely unused.
- Network Topology Flexibility: The Accurus network is designed to be aware of physical proximity, allowing devices to form local clusters for high-performance compute while still communicating globally over the internet.
3. Market/Investment Angle
- Cost Efficiency: Mobile devices offer a significantly lower cost structure for compute compared to centralized data centers due to efficient chips and the ability to acquire hardware cheaply (even second-hand/malfunctioning units). This low cost can offset the overhead associated with decentralization (e.g., replication).
- Leveraging Existing R&D: The project capitalizes on the massive R&D budgets of hardware manufacturers (like Apple’s $33B annual spend) by using their existing, highly engineered silicon without needing modification.
- New Business Models: The emergence of this infrastructure opens opportunities for businesses dealing in second-hand phones to institutionalize the process of onboarding devices into the decentralized network, mirroring the evolution of Bitcoin mining.
4. Notable Companies/People
- Alessandro Di Carli (Accurus Co-founder): The guest, whose background in mobile security informed the project’s focus on verifiable and confidential compute.
- Accurus: The project aiming to build the decentralized mobile compute network.
- Bitcoin Network: Used as the primary case study for successful, resilient, and economically incentivized decentralization.
- Node.js: Mentioned as the development framework used by Accurus to lower the barrier to entry for developers deploying workloads.
5. Regulatory/Policy Discussion
No direct regulatory discussion occurred, but the concept of confidential compute via Trusted Execution Environments (TEEs) is highlighted as a solution that allows users to run sensitive applications (like LLMs) without fear of the counterparty inspecting or tampering with the data, which has significant privacy implications.
6. Future Implications
The conversation suggests a future where decentralized compute becomes an accessible utility, similar to how Bitcoin offered an alternative financial asset. While initial adoption will likely focus on non-mission-critical tasks, the network’s global scalability and local accessibility will drive broader adoption. A key predicted use case is running confidential LLMs (like a private ChatGPT) directly on the decentralized network.
7. Target Audience
This episode is most valuable for Crypto/Web3 Infrastructure Developers, Technology Strategists, Mobile Security Professionals, and Investors interested in the convergence of decentralized systems, edge computing, and AI infrastructure.
Comprehensive Narrative Summary
The episode centers on the vision of Accurus to democratize computational resources by leveraging the billions of unused mobile phones globally. Alessandro Di Carli frames this ambition by first analyzing the Bitcoin network—the world’s most powerful decentralized computer—as proof that software rules can organize massive physical resources (miners) into an unstoppable, resilient system.
Di Carli argues that just as Bitcoin provided an alternative financial asset, decentralized compute must offer a global, accessible alternative to centralized cloud services. The critical missing piece for achieving this, he asserts, is security, specifically verifiability and confidentiality. This is where mobile phones become the ideal infrastructure. Unlike data centers, phones possess sophisticated Hardware Security Modules (HSMs) that enable Trusted Execution Environments (TEEs), ensuring that computations remain private and the work performed is verifiable, even by the device owner.
The conversation addresses the complexity of building such a network, noting that it requires not just miners and node operators (like Bitcoin), but also the initial ideologists who believe in the alternative. Furthermore, Accurus deliberately avoids partnering with hardware manufacturers, instead piggybacking on their existing R&D by utilizing the unmodified hardware and relying on standard hardware attestations to verify device authenticity.
The economic incentive is rooted in the low operational cost of mobile chips compared to traditional servers. This cost advantage is necessary to overcome the inherent overhead of decentralized systems (like running multiple replicas).
Practical applications already being tested on the Accurus testnet include data scraping (leveraging diverse consumer IP addresses) and running secure Bitcoin wallets within TEEs. The most compelling future use case highlighted is running confidential Large Language Models (LLMs), allowing users to interact with powerful AI without any third party—including the node operator—being able to read their prompts or data. The
🏢 Companies Mentioned
đź’¬ Key Insights
"Yet another example is also one protocol is utilizing this to wrap and unwrap native Bitcoins basically running a Bitcoin wallet inside the trusted execution environments and by now already holding more than 900 bitcoins on chain."
"What we offered as one of the accelerations of the environment that people can use to deploy things is basically an LLM acceleration. And what this translates to for a consumer is that you can run your own confidential ChatGPT on Accurus while you're 100% sure that the other party is not capable of reading all of the context and data that you're sharing with it."
"One thing that these devices offer you is confidential compute basically. Because of this chip that I'm now mentioning over and over again, you can basically leverage a thing called trusted execution environment that makes sure that whoever is running the device for you is incapable of inspecting whatever it is that you're giving and sharing as data."
"one of the biggest risks that we've seen now with running our testnet for over a year is the risk of someone just basically unplugging the device entirely... And the only way how you can solve that is by setting the right incentives and by basically saying, if you say that you're going to provide compute and you get paid for it, there will be a penalty if you are no longer doing what you were saying."
"mobile devices have a very nice property that whenever they have some sort of malfunction, people look at them as pretty much worthless. This means on the other side that you can with very little capital expense start buying a lot of compute, a lot of valuable compute like these devices have eight cores and 16 gigs of RAM by now for really not that much money."
"mobile devices really excel at security. They offer some of the most advanced hardware security modules out there and they are literally everywhere, right? In everybody's pocket."